Wednesday, March 14, 2012

What about autosomal testing that tells us what ethnic groups we fall into?

What about autosomal testing that tells us what ethnic groups we fall into?
There are two types of autosomal DNA tests. The first test was by DNAPrint although was marketed by several other companies under different names. It was the only test to provide percentages of ethnicity for European, African, Asian and Native American. This company has gone out of business and this test is no longer available. While initially the genetic genealogy community was very hopeful that these tests were reliable and accurate, with time and several years of experience, the results unfortunately have come to be viewed increasingly as inaccurate and unreliable for the detection of minority ancestry admixture[16]. The only people who seemed to be happy were those who received results they were seeking. Others, such as an individual from Germany whose entire family had lived there for hundreds of years, received a report that said he was 35% combined Asian and Native American. He was understandably unhappy and exceedingly skeptical[17]. While these tests are interesting and perhaps hold promise for the distant future, the technology and underlying population data bases are problematic and the tests have difficulty in detecting minority admixture accurately, tending to report higher percentages than actually exist.
The second kind of autosomal testing provides you with a list of populations or geographic locations. Two companies provide this kind of reporting based on a standard Codis autosomal test[18]. The issues with this type of testing, or more specifically the interpretation of the tests, are that the population list relies on a number of factors which are problematic. 
1. The populations are taken from forensic and medical journals and are often small studies. The population from a small village in Northern Italy, with 20 people, may not be representative of all of Italy, for example. 
2. In other cases, the population identified may be ambiguous. For example, Lumbee is a designation. What does Lumbee mean? There is not a federally recognized Lumbee tribe with blood quantum membership criteria, so who is a Lumbee? The Lumbee group is known to have been extremely admixed as early as the 1880s[19], so today, what ethnicity is a Lumbee?
3. Who identified the individuals in the study as belonging to a specific ethnic or geographic group? The individual being arrested, the booking officer, the nurse in the doctor’s office? What criteria did they use to assign that person to that group?
4. How many people were involved in a reference study? One person or a thousand people?
5. We don’t know exactly how autosomal DNA is selected to be passed from parent to child, so what exactly are we measuring and what does it really mean?
6. Brian Burritt, the forensic police officer who created OmniPop, the tool upon which both companies analysis is based[20] has gone on record stating that he created Omnipop to differentiate between people, not to find their similarities, that genealogists are using his tool for something it was not designed for and they are overanalyzing the results[21].

7. OmniPop can legitimately be run with three different sets of marker criteria, all of which are “correct”, but the results of which will be significantly different[22]. Determining which one is “right” and presented to the customer may be a function of which one best reflects what the customer is looking for in their results.
Again, satisfaction with these tools seems to be a function of how closely the results reflect the desired finding of the individual being tested.
For additional information about autosomal DNA testing in general or in relation to Native Heritage, go to and scroll down to see the various articles.

© History Chasers Click here to view all recent Historical Melungeons Blog posts Bookmark and Share

Enter your email address to start receiving this blog in your inbox

Enter your email address:
Delivered by FeedBurner